Critical design activities in house-building projects an industrial process perspective

Abstract: Dealing with planning accuracy respectively design errors is a big issue within construction. It has been argued that design errors are a major cause for waste in housing projects, despite all technical development. Reasons for design errors are well investigated and often relate to human acting. Because of the complex structure of the design process, i.e., the strong interrelation of the activities in the design phase, several activities have to be iterated if a design error at a certain activity occurs and is not detected immediately. Deviations from a planned sequence of activities, i.e., flow interruptions in Lean terminology, do not only imply rework but means also a higher risk for loosing control over the project. It can result in, e.g. poor coordination of project participants, necessary changes in schedules, possible time pressure to hold the schedule and about all a higher risk for making errors again. The long term goal of this research is to reduce cost variability of building projects by enhancing flow control in the design phase. A good flow control means to be able to carry out an activity sequence as planned. Existing literature mainly gives general, rather strategically advises what to do or not to do in order to enhance flow. But only few studies can be found dealing with the complex structure of the design process, enabling the development of concrete countermeasures at certain activities, e.g. placing checkpoints. The idea of this research is that not every faulty activity output will have the same effect on the flow in the design phase. The research question is how activities with a high impact on the flow in the design phase can be identified. A housing project was mapped in detail with focus on the activities carried out and theirs relations amongst each other in order to get a better understanding about the complexity of a building design process. The building is a two storey residential home with about 1100 m2. It is carried out in a new building system, a prefab system that is based on timber frame elements. The project was organised in an open building system and the project team consisted of around 10 contractors, small and medium sized enterprises from all over Sweden. An important postulate for this work is that, even there is a high variation of performance across projects, there is an underlying process structure, i.e. relations between the activities, which do not vary much for a certain building system. By applying design structure matrix algorithms a standard sequence of activities (reference process) has been derived. The execution of projects has been simulated via arbitrary deviations from the sequence, where a deviation had a certain consequence for the further proceeding because of the existing relations. The output variable of the simulation model is a measure for the occurrence of sequence deviations. The simulation result indicates that activities with a high number of subsequent activities with long working hours have a bigger influence on the flow in the design phase than activities that dominate only few activities with low working hours.